Could Differences in Treatment Between Trial Arms Explain the Reduction in Prostate Cancer Mortality in the European Randomized Study of Screening for Prostate Cancer?

Background: Differential treatment between trial arms has been suggested to bias prostate cancer (PC) mortality in the European Randomized Study of Screening for Prostate Cancer (ERSPC). Objective: To quantify the contribution of treatment differences to the observed PC mortality reduction between the screening arm (SA) and the control arm (CA). Design, setting, and participants: A total of 14 136 men with PC (SA: 7310; CA: 6826) in the core age group (55–69 yr) at 16 yr of follow-up. Outcome measurements and statistical analysis: The outcomes measurements were observed and estimated numbers of PC deaths by treatment allocation in the SA and CA, respectively. Primary treatment allocation was modeled using multinomial logistic regression adjusting for center, age, year, prostate-specific antigen, grade group, and tumor-node-metastasis stage. For each treatment, logistic regression models were fitted for risk of PC death, separately for the SA and CA, and using the same covariates as for the treatment allocation model. Treatment probabilities were multiplied by estimated PC death risks for each treatment based on one arm, and then summed and compared with the observed number of deaths. Results and limitations: The difference between the observed and estimated treatment distributions (hormonal therapy, radical prostatectomy, radiotherapy, and active surveillance/watchful waiting) in the two arms ranged from 3.3% to 3.3%. These figures, which represent the part of the treatment differences between arms that cannot be explained by clinicopathological differences, are small compared with the observed differences between arms that ranged between 7.2% and 10.1%. The difference between the observed and estimated numbers of PC deaths among men with PC was 0.05% (95% confidence interval [CI] 0.1%, 0.2%) when applying the CA model to the SA, had the two groups received identical primary treatment, given their clinical characteristics. When instead applying the SA model to the CA, the difference was, as expected, very

• View at Publisher